CN103730900A - Multi-time-scale province-region-county integrated reactive power optimization method of electric system - Google Patents

Abstract

The invention discloses a multi-time-scale province-region-county integrated reactive power optimization method of an electric system. The method considers the yearly rule, the season rule, the monthly rule, the day rule, super-short-time built-in inertia rules and random components. The method can achieve multi-time-scale province-region-county integrated reactive power optimization, provides multi-time-scale continuous reactive voltage optimal configuration strategies, achieves reactive voltage optimal operation between various voltage classes in long periods, intermediate periods and short periods, improves the voltage qualified rate of the whole network, guarantees continuous, safe and stable operation of the power grid, provides theoretical foundations for setting up the operation mode of the power grid and reasonably and optimally utilizing electricity generation resources, and can effectively guarantee safe, stable and continuous operation of the power grid.

Description

Province's integrated idle work optimization method in ground county of electric power system Multiple Time Scales

Technical field

The present invention relates to power system reactive power analysis and optimization field, especially relate to a kind of province's integrated idle work optimization method in ground county of efficient electric power system Multiple Time Scales.

Background technology

In recent ten years, power system safety and stability problem becomes the focus that industry is paid close attention to day by day.In modern power network, because the unify load of long transmission line of weak pattern increases the weight of, the transmission capacity of electrical network more and more approaches its limit, makes the stable problem of system more outstanding.Optimize the operational mode of electric power system, can improve significantly the safety and stability environment of electric power system, and the idle work optimization of electric power system is important step wherein, it relates to large-scale, non-linear, dynamic and discrete combinatorial problem.Existing flow process be operations staff according to load data, grid structure and experience initialization, carry out verification by the flow calculation program of BPA; Then the tidal current chart of manual analysis BPA, artificially determines each relevant adjustment amount by each point voltage and direction of tide.So, even if sophisticated operations staff, also big frame in hand, and a lot of detailed problem can not be exhaustive, find optimum scheme is very difficult thing.Along with the development of power system analysis method and computing technique, be badly in need of solving the above problems, so far there are no the many-side deficiency to traditional idle work optimization, carry out the scheme comprehensively solving.

Chinese patent mandate publication number: CN103199544A, authorize open day on July 10th, 2013, the idle work optimization method that discloses a kind of electric power system, comprises the following steps: adopt Newton-Raphson flow calculation program to calculate for optimizing value and network loss value that each state variable is provided; Set up the bacterium described reactive power optimization of power system model that optimized algorithm is combined with particle cluster algorithm of looking for food, determine optimization aim function; Initialization bacterial community, initiation parameter comprises: the probability that operator is migrated in bacterial population size, control variables number, bacterium position, chemotactic operator number of times, breeding operator number of times, the ratio of migrating operator number of times, execution breeding, execution; Continuous variable in bacterium individuality and discrete variable are carried out to initialization, the set end voltage that continuous variable is generator, discrete variable comprises: transformer gear, building-out capacitor gear; The corresponding transformer voltage ratio of transformer gear, the corresponding building-out capacitor amount of building-out capacitor gear.The weak point of this invention is not consider idle work optimization limitation between time scale and each region.

Summary of the invention

Goal of the invention of the present invention is according to line voltage grade difference, not carry out different modes idle work optimization and do not consider that the characteristic of different time yardstick electrical network carries out the deficiency of different idle work optimization modes in order to overcome in prior art, and a kind of province's integrated idle work optimization method in ground county of efficient electric power system Multiple Time Scales is provided.

To achieve these goals, the present invention is by the following technical solutions:

Province's integrated idle work optimization method in ground county of electric power system Multiple Time Scales, comprises the following steps:

(1-1) the electrical network scope that selection will be optimized, obtain the rack data of its corresponding electrical network and the service data of electrical network, in computer, be provided with several time scales, user controls computer and selects the rack data of at least one 1 time scale and service data to carry out the trend calculating of province ground county:

(1-1-1) read rack data and service data, rack data comprise Net Frame of Electric Network parameter, and service data comprises the parameter of system operation, set V _i=1 and δ _i=0, set iterations k=0, maximum iteration time k _max=M, generates admittance matrix Y according to parameter of double--layer grids _b; Admittance matrix Y _bin comprise electrical network grid structure relation;

$Y_B=\left[\begin{array}{ccccc}{Y}_{11}& Y_{12}& Y_{13}& ...& Y_{1n}\\ Y_{21}& Y_{22}& Y_{23}& ...& Y_{2n}\\ Y_{31}& Y_{32}& Y_{33}& ...& Y_{3n}\\ .& .& .& .& .\\ .& .& .& .& .\\ .& .& .& .& .\\ Y_{n1}& Y_{n2}& Y_{n3}& ...& Y_{\mathrm{nn}}\end{array}\right]$

In formula, the diagonal element Y of node admittance matrix _ii(i=1,2 ..., n) claim self-admittance; The nondiagonal element Y of node admittance matrix _ji(i=1,2 ..., n; J ≠ i) title transadmittance.

(1-1-2) utilize power flow equation $\left\{\begin{array}{c}{P}_{\mathrm{Gi}}-P_{\mathrm{Di}}-\underset{j\∈S_N}{\mathrm{\Σ}}{V}_i{Y}_{\mathrm{ij}}{V}_j\cos {\mathrm{\δ}}_{\mathrm{ij}}=0\\ Q_{\mathrm{Gi}}-Q_{\mathrm{Di}}-\underset{j\∈S_N}{\mathrm{\Σ}}{V}_i{Y}_{\mathrm{ij}}{V}_j\sin {\mathrm{\δ}}_{\mathrm{ij}}=0\end{array}\right.i\∈S_N,$ Rated output amount of unbalance Δ P _iwith Δ Q _i:

Wherein: i is any node in electrical network, j is any node outside node i in electrical network, P _gi: be the generator active power of node i; Q _gi: the generator reactive power of node i; P _di: the load active power of node i; Q _difor the reactive load power of node i; V _ifor the voltage magnitude of node i; Y _bbe the admittance matrix that the admittance unit between each node of electrical network forms, reacted the electrical structure relation of electrical network; δ _ij=δ _i-δ _j-α _ij; δ _ifor the phase angle of the voltage of node i; δ _jfor the voltage phase angle of node j; α _ijfor corresponding admittance matrix element Y _ijangle, represent the admittance angle between node i and node j; S _nfor grid nodes set; M is maximum iteration time;

(1-1-3) verification unbalanced power amount Δ P _iwith Δ Q _ivalue, if max| (Δ P _i, Δ Q _i) | < ε, convergence, then carries out Equivalent Network, wherein convergence precision ε=10 ^-6;

If max| is (Δ P _i, Δ Q _i) |>=ε, utilizes $\left[\begin{array}{cc}\frac{\&PartialD;\mathrm{\&Delta;P}}{\&PartialD;\mathrm{\&delta;}}& \frac{\&PartialD;\mathrm{\&Delta;P}}{\&PartialD;V}\\ \frac{\&PartialD;\mathrm{\&Delta;Q}}{\&PartialD;\mathrm{\&delta;}}& \frac{\&PartialD;\mathrm{\&Delta;Q}}{\&PartialD;V}\end{array}\right]\left[\begin{array}{c}\mathrm{\&Delta;\&delta;}\\ \mathrm{\&Delta;V}\end{array}\right]=-\left[\begin{array}{c}\mathrm{\&Delta;P}\\ \mathrm{\&Delta;Q}\end{array}\right]$ Calculate amount of unbalance Δ V and Δ δ;

Wherein,

for the local derviation of active power deviator to node voltage phase angle;

for the local derviation of active power deviator to voltage; for the local derviation of reactive power deviator to node voltage phase angle;

for the local derviation of reactive power deviator to voltage; Δ δ is node voltage phase angle increment; Δ V is node voltage increment, and Δ P is a dimensional vector: Δ P=(Δ P ₁, Δ P ₂... Δ P _i), i ∈ S _n; Δ Q is a dimensional vector: Δ Q=(Δ Q ₁, Δ Q ₂... Δ Q _i), i ∈ S _n;

(1-1-4) utilize equation $\left\{\begin{array}{c}\mathrm{\&Delta;V}=V^{(k+1)}-V^{\left(k\right)}\\ \mathrm{\&Delta;\&delta;}={\mathrm{\&delta;}}^{(k+1)}-{\mathrm{\&delta;}}^{\left(k\right)}\end{array}\right.$ Solve V ^(k+1)and δ ^(k+1), make k value increase by 1, as max| (Δ P _i, Δ Q _i) | during>=ε, repeating step (1-1-2);

Wherein, V ^(k+1)for iterations k increases the new voltage iterative value of all nodes of 1 rear electrical network, for

δ ^(k+1)for iterations k increases the new phase angle iterative value of all nodes of 1 rear electrical network, for ${\mathrm{\&delta;}}^{(k+1)}=({\mathrm{\&delta;}}_1^{(k+1)},{\mathrm{\&delta;}}_2^{(k+1)},\&CenterDot;\&CenterDot;\&CenterDot;{\mathrm{\&delta;}}_i^{(k+1)}),i\&Element;S_N;$

If (1-2) k > M, the active-power P to network parameter, generator by optimal load flow _gi, reactive power Q _githe size of exerting oneself is adjusted calculating:

(1-2-1) again read rack data and service data, set iterations k ₁=0, maximum iteration time is restricted to k _1max=N, arranges node voltage V _i=1 and node phase angle δ _i=0, slack variable l is set _i=1, u _i=1, Lagrange multiplier z _i=1, w _i=-0.5, y _i=0; σ ∈ (0,1) is called Center Parameter, according to parameter of double--layer grids, generates admittance matrix Y _b;

(1-2-2)

Work as k ₁during < N, calculate complementary clearance G ap=l ^tz-u ^tw, if complementary clearance G ap < ε, optimal load flow convergence, then carries out Equivalent Network, wherein convergence precision ε=10 ^-6, r is constant;

Wherein, l is a dimensional vector, l ^tthe transposition of l, l=(l ₁, l ₂... l _i), i ∈ r; U is a dimensional vector, u ^tthe transposition of u, u=(u ₁, u ₂... u _i), i ∈ r; Z is a dimensional vector: z=(z ₁, z ₂... z _i), i ∈ r; W is a dimensional vector: w=(w ₁, w ₂... w _i), i ∈ r;

If Gap>=ε, utilizes μ=σ (l ^tz-u ^tw)/(2r) calculation perturbation factor mu, and calculate update equation group $\left\{\begin{array}{c}-[{\&dtri;}_x^{2}f\left(x\right)-{\&dtri;}_x^{2}h\left(x\right)y-{\&dtri;}_x^{2}g\left(x\right)(z+w)]\mathrm{\&Delta;x}+{\&dtri;}_{x}h\left(x\right)\mathrm{\&Delta;y}+{\&dtri;}_{x}g\left(x\right)(\mathrm{\&Delta;z}+\mathrm{\&Delta;w})=L_x\\ {\&dtri;}_x{h}^T\left(x\right)\mathrm{\&Delta;x}=-L_y\\ {\&dtri;}_x{g}^T\left(x\right)\mathrm{\&Delta;x}-\mathrm{\&Delta;l}=-L_z\\ {\&dtri;}_x{g}^T\left(x\right)\mathrm{\&Delta;x}+\mathrm{\&Delta;u}=-L_w\\ \mathrm{Z\&Delta;l}+\mathrm{L\&Delta;z}=-L_l^{\mathrm{\&mu;}}\\ \mathrm{W\&Delta;u}+\mathrm{U\&Delta;w}=-L_u^{\mathrm{\&mu;}}\end{array}\right.,$

$\left\{\begin{array}{c}{L}_x\&equiv;{\&dtri;}_{x}f\left(x\right)-{\&dtri;}_{x}h\left(x\right)y-{\&dtri;}_{x}g\left(x\right)(z+w)=0\\ L_y\&equiv;h\left(x\right)=0\\ L_z\&equiv;g\left(x\right)-l-\underset{\&OverBar;}{g}=0\\ L_w\&equiv;g\left(x\right)+u-\stackrel{\&OverBar;}{g}=0\end{array}\right.;$

Wherein, disturbance complementarity condition be $\left\{\begin{array}{c}{L}_l^{\mathrm{\&mu;}}=\mathrm{LZe}-\mathrm{\&mu;e}=0\\ L_u^{\mathrm{\&mu;}}=\mathrm{UWe}+\mathrm{\&mu;e}=0\end{array}\right.,$

Target function: $f(\&CenterDot;)=\underset{i\&Element;S_{\mathrm{GN}}}{\mathrm{\&Sigma;}}{P}_{\mathrm{Gi}}-\underset{i\&Element;S_{\mathrm{DN}}}{\mathrm{\&Sigma;}}{P}_{\mathrm{Di}}$

Variable: x=(P _g, Q _r, δ, V), wherein, h (x) is power flow equation (1) formula; ▽ _xh (x) is the single order local derviation of power flow equation to state variable;

for the second order local derviation of power flow equation to state variable; G (x) is inequation group; ▽ _xg (x) is the single order local derviation of inequation group to state variable;

for the second order local derviation of inequation group to state variable; L represents with l _ifor cornerwise matrix; Z represents with z _ifor cornerwise matrix; W represents with w _icornerwise matrix; ▽ _xf (x) is the first derivative of target function to state variable; the second dervative of target function to state variable; P _gi: i platform unit is meritorious exerts oneself; P _di: the burden with power of i node; S _gN: generator set; S _dN: load aggregation; One dimensional vector of e representation unit 1; N is maximum iteration time, according to the moved determined limited cycle-index of province ground county electrical network size; X=(P _g, Q _r, δ, V), P _grepresent the control variables of generator node; Q _rrepresent the control variables of PV node; δ represents the state variable of the voltage phase angle of all nodes; V represents the voltage magnitude state variable of all nodes;

(1-2-3) calculate Δ x, Δ y, Δ l, Δ u, Δ z, Δ w, utilize $\left\{\begin{array}{c}{\mathrm{\&alpha;}}_p=0.9995\min \{\underset{i}{\min }(\frac{-l_i}{\mathrm{\&Delta;}{l}_i},\mathrm{\&Delta;}{l}_i<0;\frac{-u_i}{\mathrm{\&Delta;}{u}_i},\mathrm{\&Delta;}{u}_i<0),1\}\\ {\mathrm{\&alpha;}}_d=0.9995\min \{\underset{i}{\min }(\frac{-z_i}{\mathrm{\&Delta;}{z}_i},\mathrm{\&Delta;}{z}_i<0;\frac{-w_i}{\mathrm{\&Delta;}{w}_i},\mathrm{\&Delta;}{w}_i>0),1\}\end{array}\right.i=\mathrm{1,2},...,r,$ Calculate iteration step length α _pand α _d;

(1-2-4) utilize $\left\{\begin{array}{c}{x}^{(k+1)}=x^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;x}\\ l^{(k+1)}=l^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;l}\\ u^{(k+1)}=u^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;u}\end{array}\right.$ With $\left\{\begin{array}{c}{y}^{(k+1)}=y^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;y}\\ z_l^{(k+1)}=z^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;z}\\ w^{(k+1)}=w^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;w}\end{array}\right.$ Calculate x ^(k+1), l ^(k+1), u ^(k+1), y ^(k+1),

and w ^(k+1); Make k ₁value increase by 1, proceed to step (1-2-1);

If (1-3) max| (Δ P _i, Δ Q _i) | < ε, Ze Jiangshengdi county data file is carried out Ward Equivalent Network, obtains rack data and system service data;

(1-4) whether repeating step (1-1-1) to (1-1-4) calculating rack data and system service data restrain, if k ₁> N, user controls computer and selects the rack data of yardstick At All Other Times and system service data repeating step (1-1-1) to (1-3) to carry out province ground county trend to calculate;

If (1-5) trend is calculated max| (Δ P _i, Δ Q _i) | < ε parses required rack data and the system service data of the province ground integrated idle work optimization in county of Multiple Time Scales from the equivalent network data file generating;

(1-6) province's required rack data and service data of the ground integrated idle work optimization in county of the Multiple Time Scales parsing carried out to combined optimization:

(1-6-1) read rack data and service data, set iterations k ₂=0, maximum iteration time is restricted to k _max=M2, sets node voltage V _i=1 and node phase angle δ _i=0, set slack variable l _i=1, u _i=1, Lagrange multiplier z _i=1, w _i=-0.5, y _i=0; Wherein, Center Parameter σ ∈ (0,1), generates admittance matrix Y according to parameter of double--layer grids _b;

(1-6-2) work as k ₂during < M2, calculate complementary clearance G ap=l ^tz-u ^tw, if complementary clearance G ap < ε, OPTIMAL REACTIVE POWER is calculated convergence, generates the destination file that comprises each node reactive power distribution of electrical network, transformer station's reactive apparatus switching quantity and load tap changer position, wherein convergence precision ε=10 ^-6; If Gap>=ε, utilizes μ=σ (l ^tz-u ^tw)/(2r) calculation perturbation factor mu, and calculate update equation group

$\left\{\begin{array}{c}-[{\&dtri;}_x^{2}f\left(x\right)-{\&dtri;}_x^{2}h\left(x\right)y-{\&dtri;}_x^{2}g\left(x\right)(z+w)]\mathrm{\&Delta;x}+{\&dtri;}_{x}h\left(x\right)\mathrm{\&Delta;y}+{\&dtri;}_{x}g\left(x\right)(\mathrm{\&Delta;z}+\mathrm{\&Delta;w})=L_x\\ {\&dtri;}_x{h}^T\left(x\right)\mathrm{\&Delta;x}=-L_y\\ {\&dtri;}_x{g}^T\left(x\right)\mathrm{\&Delta;x}-\mathrm{\&Delta;l}=-L_z\\ {\&dtri;}_x{g}^T\left(x\right)\mathrm{\&Delta;x}+\mathrm{\&Delta;u}=-L_w\\ \mathrm{Z\&Delta;l}+\mathrm{L\&Delta;z}=-L_l^{\mathrm{\&mu;}}\\ \mathrm{W\&Delta;u}+\mathrm{U\&Delta;w}=-L_u^{\mathrm{\&mu;}}\end{array}\right.;$

$\left\{\begin{array}{c}{L}_x\&equiv;{\&dtri;}_{x}f\left(x\right)-{\&dtri;}_{x}h\left(x\right)y-{\&dtri;}_{x}g\left(x\right)(z+w)=0\\ L_y\&equiv;h\left(x\right)=0\\ L_z\&equiv;g\left(x\right)-l-\underset{\&OverBar;}{g}=0\\ L_w\&equiv;g\left(x\right)+u-\stackrel{\&OverBar;}{g}=0\end{array}\right.$

Disturbance complementarity condition $\left\{\begin{array}{c}{L}_l^{\mathrm{\&mu;}}=\mathrm{LZe}-\mathrm{\&mu;e}=0\\ L_u^{\mathrm{\&mu;}}=\mathrm{UWe}+\mathrm{\&mu;e}=0\end{array}\right.,$

$P_{\mathrm{ij}}^{\mathrm{set}}={\left(V_i^{\mathrm{set}}\right)}^2{Y}_{\mathrm{ij}}^{\mathrm{set}}\cos {\mathrm{\&alpha;}}_{\mathrm{ij}}^{\mathrm{set}}-V_i^{\mathrm{set}}{Y}_{\mathrm{ij}}^{\mathrm{set}}{V}_j^{\mathrm{set}}\cos {\mathrm{\&delta;}}_{\mathrm{ij}}^{\mathrm{set}},$

${\left(I_{\mathrm{ij}}^{\mathrm{set}}\right)}^2={\left(V_i^{\mathrm{set}}\right)}^2{\left(Y_{\mathrm{ij}}^{\mathrm{set}}\right)}^2+{\left(V_j^{\mathrm{set}}\right)}^2{\left(Y_{\mathrm{ij}}^{\mathrm{set}}\right)}^2-2V_i^{\mathrm{set}}{\left(Y_{\mathrm{ij}}^{\mathrm{set}}\right)}^2{V}_j^{\mathrm{set}}\cos ({\mathrm{\&delta;}}_i^{\mathrm{set}}+{\mathrm{\&alpha;}}_{\mathrm{ij}}^{\mathrm{set}})\cos ({\mathrm{\&delta;}}_j^{\mathrm{set}}+{\mathrm{\&alpha;}}_{\mathrm{ij}}^{\mathrm{set}})$

$-2V_i^{\mathrm{set}}{\left(Y_{\mathrm{ij}}^{\mathrm{set}}\right)}^2{V}_j^{\mathrm{set}}\sin ({\mathrm{\&delta;}}_i^{\mathrm{set}}+{\mathrm{\&alpha;}}_{\mathrm{ij}}^{\mathrm{set}})\sin ({\mathrm{\&delta;}}_j^{\mathrm{set}}+{\mathrm{\&alpha;}}_{\mathrm{ij}}^{\mathrm{set}})$

Wherein, the number that r is inequality constraints,

$h\left(x\right)=\left\{\begin{array}{cc}{P}_{\mathrm{Gi}}^{\mathrm{set}}-P_{\mathrm{Di}}^{\mathrm{set}}-\underset{j\&Element;S_N^{\mathrm{set}}}{\mathrm{\&Sigma;}}{V}_i^{\mathrm{set}}{Y}_{\mathrm{ij}}^{\mathrm{set}}{V}_j^{\mathrm{set}}\cos {\mathrm{\&delta;}}_{\mathrm{ij}}^{\mathrm{set}}=0& i\&Element;S_N^{\mathrm{set}}\\ Q_{\mathrm{Gi}}^{\mathrm{set}}-Q_{\mathrm{Di}}^{\mathrm{set}}-\underset{j\&Element;S_N^{\mathrm{set}}}{\mathrm{\&Sigma;}}{V}_i^{\mathrm{set}}{Y}_{\mathrm{ij}}^{\mathrm{set}}{V}_j^{\mathrm{set}}\sin {\mathrm{\&delta;}}_{\mathrm{ij}}^{\mathrm{set}}=0& i\&Element;S_N^{\mathrm{set}}\\ (k_{\mathrm{ij}}^{\mathrm{set}1}-k_{\mathrm{ij}1}^{\mathrm{set}1})(k_{\mathrm{ij}}^{\mathrm{set}1}-k_{\mathrm{ij}2}^{\mathrm{set}1})\&CenterDot;\&CenterDot;\&CenterDot;(k_{\mathrm{ij}}^{\mathrm{set}1}-k_{\mathrm{ijm}}^{\mathrm{set}1})=0& i,j\&Element;S_K^{\mathrm{set}1}\\ (Q_{\mathrm{Ci}}^{\mathrm{set}}-Q_{\mathrm{Ci}1}^{\mathrm{set}})(Q_{\mathrm{Ci}}^{\mathrm{set}}-Q_{\mathrm{Ci}2}^{\mathrm{set}})\&CenterDot;\&CenterDot;\&CenterDot;(Q_{\mathrm{Ci}}^{\mathrm{set}}-Q_{\mathrm{Cim}}^{\mathrm{set}})=0& i\&Element;S_T^{\mathrm{set}}\end{array}\right.$

Target function: $f(\&CenterDot;)=\left[\begin{array}{c}\underset{i\&Element;S_{\mathrm{GN}}^{\mathrm{set}}}{\mathrm{\&Sigma;}}{P}_{\mathrm{Gi}}^{\mathrm{set}}-\underset{i\&Element;S_{\mathrm{DN}}^{\mathrm{set}}}{\mathrm{\&Sigma;}}{P}_{\mathrm{Di}}^{\mathrm{set}}\\ \underset{i\&Element;S_{\mathrm{GN}}^{\mathrm{set}}}{\mathrm{\&Sigma;}}{Q}_{\mathrm{Gi}}^{\mathrm{set}}-\underset{i\&Element;S_{\mathrm{DN}}^{\mathrm{set}}}{\mathrm{\&Sigma;}}{Q}_{\mathrm{Di}}^{\mathrm{set}}\end{array}\right]$

Variable: x=(P _g, Q _r, δ, V)

Wherein, h (x) is power flow equation formula; ▽ _xh (x) is the single order local derviation of power flow equation to state variable;

for the second order local derviation of power flow equation to state variable; G (x) is inequation group; ▽ _xg (x) is the single order local derviation of inequation group to state variable;

for the second order local derviation of inequation group to state variable; L represents with l _ifor cornerwise matrix; Z represents with z _ifor cornerwise matrix; W represents with w _icornerwise matrix; ▽ _xf (x) is the first derivative of target function to state variable;

the second dervative of target function to state variable; α _ij: the angle of node admittance matrix corresponding element; meritorious the exerting oneself of province's ground county i platform unit;

province's ground county i node burden with power;

idle the exerting oneself of province's ground county i platform unit;

province's ground county i node load or burden without work;

province's ground county's generator set;

province ground county load aggregation; for economize each gear respective value of this transformer tapping, m is its total gear,

for economizing ground adjustable transformer group set;

for idle value of exerting oneself of correspondence of this tunable capacitor group switching group number, m is its total group number,

for province's ground county's tunable capacitor group set; One dimensional vector of e representation unit 1; M2 is maximum iteration time, according to the moved determined limited cycle-index of province ground county electrical network size; { implication of d} is respectively subscript set ∈ for s, x: s is for economizing, and d is ground, and x is county; Because transformer at county level is non-adjustable, therefore voltage regulator tap-c hange control is only for the transformer under the ground of provinceing, subscript set1 ∈ s, and the implication of d} is respectively: s is for economizing, and d is; X=(P _g, Q _r, δ, V), P _grepresent the control variables of generator node; Q _rrepresent the control variables of PV node; δ represents the state variable of the voltage phase angle of all nodes; V represents the voltage magnitude state variable of all nodes;

(1-6-3) calculate Δ x, Δ y, Δ l, Δ u, Δ z, Δ w, utilize $\left\{\begin{array}{c}{\mathrm{\&alpha;}}_p=0.9995\min \{\underset{i}{\min }(\frac{-l_i}{\mathrm{\&Delta;}{l}_i},\mathrm{\&Delta;}{l}_i<0;\frac{-u_i}{\mathrm{\&Delta;}{u}_i},\mathrm{\&Delta;}{u}_i<0),1\}\\ {\mathrm{\&alpha;}}_d=0.9995\min \{\underset{i}{\min }(\frac{-z_i}{\mathrm{\&Delta;}{z}_i},\mathrm{\&Delta;}{z}_i<0;\frac{-w_i}{\mathrm{\&Delta;}{w}_i},\mathrm{\&Delta;}{w}_i>0),1\}\end{array}\right.i=\mathrm{1,2},...,r,$ Calculate iteration step length α _pand α _d;

(1-6-4) utilize $\left\{\begin{array}{c}{x}^{(k+1)}=x^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;x}\\ l^{(k+1)}=l^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;l}\\ u^{(k+1)}=u^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;u}\end{array}\right.$ With $\left\{\begin{array}{c}{y}^{(k+1)}=y^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;y}\\ z_l^{(k+1)}=z^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;z}\\ w^{(k+1)}=w^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;w}\end{array}\right.$ Calculate x ^(k+1), l ^(k+1), u ^(k+1), y ^(k+1),

and w ^(k+1); Make k ₂value increases by 1, proceeds to step (1-6-2);

If (1-7) province of the Multiple Time Scales integrated idle work optimization in ground county calculates complementary clearance G ap < ε, idle work optimization convergence, calculates to the adjusting of main transformer tap changer gear, reactive-load compensator switching amount and part transformer station increase capacitor and reactor details is upgraded initial data file and generate Xin Shengdi county data file according to idle work optimization;

If (1-8) iterations k ₂> M2 means that the integrated idle work optimization in province ground county of Multiple Time Scales calculates the province ground county data file check results not restraining or generate and do not restrain, according to slack variable, adjust inequality constraints bound, proceed to step (1-6-1);

(1-9) according to idle work optimization module, calculate the destination file generating after convergence, destination file comprises that the reactive power of whole system distributes, the reactive power compensation amount of each node, transformer voltage ratio tap joint position, reactive power compensator switching amount, node voltage distribute, reactive power exchange amount, the reactive power exchange amount between Di Yu county, the meritorious of power plant between reactive loss, province and the ground of the active loss of system, system are exerted oneself and idle exerting oneself, to the classify electrical network analysis form of formation standard of data;

(1-10) computer is adjusted strategy report according to the analysis result in electrical network analysis form and operational mode, regulate main transformer tap changer gear, regulate each reactive-load compensator switching amount in electrical network and regulate part transformer station in electrical network capacitor and or the quantity of reactor, determine local generator terminal voltage and the power of generator node in the transformer tapping gear, capacitor switching group number, electrical network of Mei Zuo transformer station.

The integrated idle work optimization method in province ground county of Multiple Time Scales of the present invention be a kind ofly considered year rule, season rule, month rule, day rule, the province integrated idle work optimization method in ground county of built-in inertia rule and random composition when ultrashort.Can realize province's integrated idle work optimization in ground county of Multiple Time Scales, provide the continuous reactive Voltage Optimum collocation strategy of Multiple Time Scales, realize the reactive Voltage Optimum operation between each electric pressure of electric power system long, medium and short phase, improve the rate of qualified voltage of the whole network, ensure the lasting safe and stable operation of electrical network, for formulating the operational mode of electrical network, rationally and the utilization of optimizing generating Resource Supply theoretical foundation, can effectively guarantee electricity net safety stable continuous service.

As preferably, time scale is 5, be respectively year, season, monthly, day, rolling ultra-short term; Rolling ultra-short term is 15 minutes.

As preferably, described rack packet is containing the network line parameter of province Di Xian whole distract or subregion, ground connection branch road parameter and transformer parameter, and network line parameter comprises reactance value x _lineij, resistance value r _lineijwith admittance value b _lineij, ground connection branch road parameter comprises electric conductivity value G _iwith susceptance value B _i, transformer parameter comprises reactance value x _lineij, resistance value r _lineijwith no-load voltage ratio k _lineij;

System operational parameters module comprises province's ground whole network in county or the operational factor of subregion and PV node parameter, and operational factor comprises generator node operational factor P _gi, Q _gioperational factor P with load bus _di, Q _di, PV node parameter comprises node voltage V _iwith reactive power bound q _i,

As preferably, M is 3 to 20, N to be 3 to 20, M2 to be 10 to 100.

As preferably, destination file comprises that the reactive power of whole system distributes, the reactive power compensation amount of each node, transformer voltage ratio tap joint position, reactive power compensator switching amount, node voltage distribute, reactive power exchange amount, the reactive power exchange amount between Di Yu county, the meritorious of power plant between reactive loss, province and the ground of the active loss of system, system are exerted oneself and idle exerting oneself.

Compared with prior art scheme, the invention has the beneficial effects as follows:

For current reactive power optimization management disperse, numerous and diverse feature, the integrated idle work optimization in province ground county of having designed and Implemented a kind of advanced person's Multiple Time Scales calculates the method for putting, computational process can be calculated by the whole network simultaneous, also can pass through boundary condition parallel computation, and account form is flexible.The method is integrated in one high, normal, basic province ground county network idle work optimization calculating, analysis, utilizes idle work optimization calculating, comparison, the analysis means of science, for realizing production decision, provides technical support.Method mainly divides Three Estate to carry out idle work optimization, is mainly: idle work optimization, classification idle work optimization, overall idle work optimization separately.Idle work optimization is followed the tracks of compensation for load level and the quality of voltage in somewhere separately, realizes local compensation and in-situ balancing, reduces line loss, improves compensation efficiency.Classification idle work optimization for different barrier electric pressures and customer charge type, needs the mode difference of reactive power compensation.Overall situation idle work optimization is for independent idle work optimization and the unreasonable scheme of classification idle work optimization, to cause that local idle drug on the market or the problem such as overall idle imbalance is carried out idle work optimization, add the method and also consider multi-period time scale, the electrical network characteristic rule that more can consider to adequacy electrical network complexity is carried out idle work optimization, can make whole distract reactive apparatus optimal utilization, System Reactive Power reaches balance, reactive-load compensation equipment expense reduces, also reduced the via net loss of electrical network simultaneously, guarantee the quality of voltage between each electric pressure, meet relevant laws and regulations of the state Its Relevant Technology Standards.Solved the many-side deficiency of traditional idle work optimization, all beyond example at home and abroad.

Accompanying drawing explanation

Fig. 1 is a kind of flow chart of embodiments of the invention.

Embodiment

With reference to the accompanying drawings, describe specific embodiment of the invention scheme in detail.

Embodiment is a kind of province's integrated idle work optimization method in ground county of combining province's ground county's electric power system Multiple Time Scales as shown in Figure 1, comprises the steps:

Step 100 is selected the electrical network scope that will optimize, the rack data of corresponding its electrical network and the service data of electrical network, in computer, be provided with 2011 years, summer in 2011, in May, 2011, on May 15th, 2011, user controls rack data and the service data that computer select time yardstick is 2011 years and carries out the trend calculating of province ground county:

Step 101, reads rack data and service data, and rack data comprise Net Frame of Electric Network parameter, and service data comprises the parameter of system operation, sets V _i=1 and δ _i=0, set iterations k=0, maximum iteration time k _max=M, generates admittance matrix Y according to parameter of double--layer grids _b; Admittance matrix Y _bin comprise electrical network grid structure relation;

Step 102, utilizes power flow equation, rated output amount of unbalance Δ P _iwith Δ Q _i:

Step 103, verification unbalanced power amount Δ P _iwith Δ Q _ivalue, if max| (Δ P _i, Δ Q _i) | < ε, convergence, then carries out Equivalent Network, wherein convergence precision ε=10 ^-6;

If max| is (Δ P _i, Δ Q _i) |>=ε utilizes $\left[\begin{array}{cc}\frac{\&PartialD;\mathrm{\&Delta;P}}{\&PartialD;\mathrm{\&delta;}}& \frac{\&PartialD;\mathrm{\&Delta;P}}{\&PartialD;V}\\ \frac{\&PartialD;\mathrm{\&Delta;Q}}{\&PartialD;\mathrm{\&delta;}}& \frac{\&PartialD;\mathrm{\&Delta;Q}}{\&PartialD;V}\end{array}\right]\left[\begin{array}{c}\mathrm{\&Delta;\&delta;}\\ \mathrm{\&Delta;V}\end{array}\right]=-\left[\begin{array}{c}\mathrm{\&Delta;P}\\ \mathrm{\&Delta;Q}\end{array}\right]$ Calculate amount of unbalance Δ V and Δ δ;

Step 104, utilizes $\left\{\begin{array}{c}\mathrm{\&Delta;V}=V^{(k+1)}-V^{\left(k\right)}\\ \mathrm{\&Delta;\&delta;}={\mathrm{\&delta;}}^{(k+1)}-{\mathrm{\&delta;}}^{\left(k\right)}\end{array}\right.$ Solve V ^(k+1)and δ ^(k+1), make k value increase by 1, as max| (Δ P _i, Δ Q _i) | during>=ε, repeating step 102;

Step 200, if k > 20, the active-power P to network parameter, generator by optimal load flow _gi, reactive power Q _githe size of exerting oneself is adjusted calculating:

Step 201, reads rack data and service data again, sets iterations k ₁=0, maximum iteration time is restricted to k _1max=20, node voltage V is set _i=1 and node phase angle δ _i=0, slack variable l is set _i=1, u _i=1, Lagrange multiplier z _i=1, w _i=-0.5, y _i=0; σ ∈ (0,1) is called Center Parameter, according to parameter of double--layer grids, generates admittance matrix Y _b;

Step 202, works as k ₁during < 20, calculate complementary clearance G ap=l ^tz-u ^tw, if complementary clearance G ap < ε, optimal load flow convergence, then carries out Equivalent Network, wherein convergence precision ε=10 ^-6;

If Gap>=ε, utilizes μ=σ (l ^tz-u ^tw)/(2r) calculation perturbation factor mu, and calculate update equation group $\left\{\begin{array}{c}-[{\&dtri;}_x^{2}f\left(x\right)-{\&dtri;}_x^{2}h\left(x\right)y-{\&dtri;}_x^{2}g\left(x\right)(z+w)]\mathrm{\&Delta;x}+{\&dtri;}_{x}h\left(x\right)\mathrm{\&Delta;y}+{\&dtri;}_{x}g\left(x\right)(\mathrm{\&Delta;z}+\mathrm{\&Delta;w})=L_x\\ {\&dtri;}_x{h}^T\left(x\right)\mathrm{\&Delta;x}=-L_y\\ {\&dtri;}_x{g}^T\left(x\right)\mathrm{\&Delta;x}-\mathrm{\&Delta;l}=-L_z\\ {\&dtri;}_x{g}^T\left(x\right)\mathrm{\&Delta;x}+\mathrm{\&Delta;u}=-L_w\\ \mathrm{Z\&Delta;l}+\mathrm{L\&Delta;z}=-L_l^{\mathrm{\&mu;}}\\ \mathrm{W\&Delta;u}+\mathrm{U\&Delta;w}=-L_u^{\mathrm{\&mu;}}\end{array}\right.,$

$\left\{\begin{array}{c}{L}_x\&equiv;{\&dtri;}_{x}f\left(x\right)-{\&dtri;}_{x}h\left(x\right)y-{\&dtri;}_{x}g\left(x\right)(z+w)=0\\ L_y\&equiv;h\left(x\right)=0\\ L_z\&equiv;g\left(x\right)-l-\underset{\&OverBar;}{g}=0\\ L_w\&equiv;g\left(x\right)+u-\stackrel{\&OverBar;}{g}=0\end{array}\right.;$

Step 203, calculates Δ x, Δ y, Δ l, Δ u, Δ z, Δ w, utilizes equation $\left\{\begin{array}{c}{\mathrm{\&alpha;}}_p=0.9995\min \{\underset{i}{\min }(\frac{-l_i}{\mathrm{\&Delta;}{l}_i},\mathrm{\&Delta;}{l}_i<0;\frac{-u_i}{\mathrm{\&Delta;}{u}_i},\mathrm{\&Delta;}{u}_i<0),1\}\\ {\mathrm{\&alpha;}}_d=0.9995\min \{\underset{i}{\min }(\frac{-z_i}{\mathrm{\&Delta;}{z}_i},\mathrm{\&Delta;}{z}_i<0;\frac{-w_i}{\mathrm{\&Delta;}{w}_i},\mathrm{\&Delta;}{w}_i>0),1\}\end{array}\right.i=\mathrm{1,2},...,r,$ Calculate iteration step length α _pand α _d;

Step 204, utilizes $\left\{\begin{array}{c}{x}^{(k+1)}=x^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;x}\\ l^{(k+1)}=l^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;l}\\ u^{(k+1)}=u^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;u}\end{array}\right.$ With $\left\{\begin{array}{c}{y}^{(k+1)}=y^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;y}\\ z_l^{(k+1)}=z^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;z}\\ w^{(k+1)}=w^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;w}\end{array}\right.$ Calculate x ^(k+1), l ^(k+1), u ^(k+1), y ^(k+1), and w ^(k+1); Make k ₁value increase by 1, proceed to step 201;

Step 300, if max| is (Δ P _i, Δ Q _i) | < ε, Ze Jiangshengdi county data file is carried out Ward Equivalent Network, obtains rack data and system service data;

Step 400, whether repeating step 101 to 104 calculating rack data and system service data restrain, if k ₁> N, user controls computer and selects the rack data of yardstick At All Other Times and system service data repeating step 101 to 300 to carry out province ground county trend to calculate;

Step 500, if trend is calculated max| (Δ P _i, Δ Q _i) | < ε parses required rack data and the system service data of the province ground integrated idle work optimization in county of Multiple Time Scales from the equivalent network data file generating;

Step 600, province's required rack data and service data of the ground integrated idle work optimization in county of the Multiple Time Scales parsing carried out to combined optimization:

Step 601, combines according to the grid structure parameter of gained and system operational parameters transformer, capacitor, the reactor parameter given with database, generates the needed data file of province's ground integrated idle work optimization computing module in county of Multiple Time Scales;

Step 602, reads rack data and service data, sets iterations k ₂=0, maximum iteration time is restricted to k _max=50, set node voltage V _i=1 and node phase angle δ _i=0, set slack variable l _i=1, u _i=1, Lagrange multiplier z _i=1, w _i=-0.5, y _i=0; Wherein, Center Parameter σ ∈ (0,1), generates admittance matrix Y according to parameter of double--layer grids _b;

Step 603, works as k ₂during < 50, calculate complementary clearance G ap=l ^tz-u ^tw, if complementary clearance G ap < ε, OPTIMAL REACTIVE POWER is calculated convergence, generates the destination file that comprises each node reactive power distribution of electrical network, transformer station's reactive apparatus switching quantity and load tap changer position, wherein convergence precision ε=10 ^-6; If Gap>=ε, utilizes μ=σ (l ^tz-u ^tw)/(2r) calculation perturbation factor mu, and calculate update equation group

$\left\{\begin{array}{c}-[{\&dtri;}_x^{2}f\left(x\right)-{\&dtri;}_x^{2}h\left(x\right)y-{\&dtri;}_x^{2}g\left(x\right)(z+w)]\mathrm{\&Delta;x}+{\&dtri;}_{x}h\left(x\right)\mathrm{\&Delta;y}+{\&dtri;}_{x}g\left(x\right)(\mathrm{\&Delta;z}+\mathrm{\&Delta;w})=L_x\\ {\&dtri;}_x{h}^T\left(x\right)\mathrm{\&Delta;x}=-L_y\\ {\&dtri;}_x{g}^T\left(x\right)\mathrm{\&Delta;x}-\mathrm{\&Delta;l}=-L_z\\ {\&dtri;}_x{g}^T\left(x\right)\mathrm{\&Delta;x}+\mathrm{\&Delta;u}=-L_w\\ \mathrm{Z\&Delta;l}+\mathrm{L\&Delta;z}=-L_l^{\mathrm{\&mu;}}\\ \mathrm{W\&Delta;u}+\mathrm{U\&Delta;w}=-L_u^{\mathrm{\&mu;}}\end{array}\right.;$

$\left\{\begin{array}{c}{L}_x\&equiv;{\&dtri;}_{x}f\left(x\right)-{\&dtri;}_{x}h\left(x\right)y-{\&dtri;}_{x}g\left(x\right)(z+w)=0\\ L_y\&equiv;h\left(x\right)=0\\ L_z\&equiv;g\left(x\right)-l-\underset{\&OverBar;}{g}=0\\ L_w\&equiv;g\left(x\right)+u-\stackrel{\&OverBar;}{g}=0\end{array}\right.;$

Step 604, calculates Δ x, Δ y, Δ l, Δ u, Δ z, Δ w, utilizes equation $\left\{\begin{array}{c}{\mathrm{\&alpha;}}_p=0.9995\min \{\underset{i}{\min }(\frac{-l_i}{\mathrm{\&Delta;}{l}_i},\mathrm{\&Delta;}{l}_i<0;\frac{-u_i}{\mathrm{\&Delta;}{u}_i},\mathrm{\&Delta;}{u}_i<0),1\}\\ {\mathrm{\&alpha;}}_d=0.9995\min \{\underset{i}{\min }(\frac{-z_i}{\mathrm{\&Delta;}{z}_i},\mathrm{\&Delta;}{z}_i<0;\frac{-w_i}{\mathrm{\&Delta;}{w}_i},\mathrm{\&Delta;}{w}_i>0),1\}\end{array}\right.i=\mathrm{1,2},...,r,$ Calculate iteration step length α _pand α _d;

Step 605, utilizes $\left\{\begin{array}{c}{x}^{(k+1)}=x^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;x}\\ l^{(k+1)}=l^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;l}\\ u^{(k+1)}=u^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;u}\end{array}\right.$ With $\left\{\begin{array}{c}{y}^{(k+1)}=y^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;y}\\ z_l^{(k+1)}=z^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;z}\\ w^{(k+1)}=w^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;w}\end{array}\right.$ Calculate x ^(k+1), l ^(k+1), u ^(k+1), y ^(k+1), and w ^(k+1); Make k ₂value increases by 1, proceeds to step 603;

Step 700, if the province of the Multiple Time Scales integrated idle work optimization in ground county calculates complementary clearance G ap < ε, idle work optimization convergence, calculates to the adjusting of main transformer tap changer gear, reactive-load compensator switching amount and part transformer station increase capacitor and reactor details is upgraded initial data file and generate Xin Shengdi county data file according to idle work optimization;

Step 800, if iterations k ₂> 50 means that the integrated idle work optimization in province ground county of Multiple Time Scales calculates the province ground county data file check results not restraining or generate and do not restrain, and adjusts inequality constraints bound, proceeds to step 602;

Step 900, calculates the destination file generating after convergence according to idle work optimization module, to the classify electrical network analysis form of formation standard of data;

Step 1000, computer is adjusted strategy report according to the analysis result in electrical network analysis form and operational mode, regulate main transformer tap changer gear, regulate each reactive-load compensator switching amount in electrical network and regulate part transformer station in electrical network capacitor and or the quantity of reactor, determine local generator terminal voltage and the power of generator node in the transformer tapping gear, capacitor switching group number, electrical network of Mei Zuo transformer station.

Step 1100, the state value of the reactive power/voltage control equipment before the integrated idle work optimization computing module calculating in province ground county of Multiple Time Scales and after calculating is added up and formed reactive voltage analysis report, comprising: network loss statistical conditions before and after 500kV transformer tapping position and corresponding voltage value, 500kV reactive compensation capacity of substation summary sheet, 220kV reactive compensation capacity of substation summary sheet, electrical network tracking unit output summary sheet, optimization; Network loss statistical conditions before and after 110kV reactive compensation capacity of substation summary sheet, electrical network tracking unit output summary sheet, optimization; Network loss statistical conditions before and after 35kV reactive compensation capacity of substation summary sheet, optimization; Network loss statistical conditions before and after 10kV optimizes;

Step 1200, by analyzing the result of calculation of the integrated idle work optimization computing module in province ground county of Multiple Time Scales, provide the optimisation strategy of the electrical network province idle unified collocation in ground county, for instructing the adjustment of actual operating, comprise: the each electric pressure tap of transformer gear is adjusted, capacitor under current reactive power compensation configuring condition and reactor switching amount, part transformer station increases the suggestion of capacitor and reactor configuration, through the adjustment of these strategies, can make active power loss and the reactive power loss of electrical network reduce, the quality of voltage of electrical network improves, the stability of system is strengthened.

It is the in the situation that of given system network architecture and system loading that the integrated idle work optimization in province ground county of Multiple Time Scales of the present invention calculates, assurance meets under the prerequisite of user power utilization demand, optimize the static service conditions of electric power system, by regulating control variables to make the target function value of appointment reach minimum, meet physical restriction and the operation restriction of system to control variables, state variable and variable function simultaneously.Adjusting control variables comprises meritorious idle the exerting oneself of generator, load tap changer, capacitor group grouping switching etc.By idle work optimization, not only can guarantee the quality of voltage of the whole network, and can obtain considerable economic benefit, can realize fail safe and the economy of the quality of power supply, system operation.

Control variables parameter setting in the present invention comprises: meritorious exert oneself, idlely exert oneself, transformer tapping, capacity reactance device switching group number.Wherein: meritorious exert oneself be set to as requested by some power plant meritorious exert oneself be made as adjustable; Idle exerting oneself requires all balance node idle of the value nodes such as meritorious conditional PV node, electrical network be outer and electrical network to exert oneself all adjustable; Transformer tapping is all adjustable; Electrical network 500kV, 220kV, 110kV, 35kV, 10kV transformer station capacity reactance device is all adjustable.The maximum permission of the capacitor compensation capacity of all transformer stations is 80% of existing configuration compensation capacity, and the maximum compensation capacity that allows of reactor of all transformer stations is existing reactor configuration capacity.

Its constraints parameter setting comprises: load or burden without work model, node voltage.Wherein: load or burden without work model all adds constrained; Node voltage restriction comprises 500kV, 220kV, 110kV, 35kV, 10kV busbar voltage, generator voltage, and limited field is as shown in table 1.

Table 1 node voltage limited field:

Certain electrical network 500kV main transformer tap joint position before and after the integrated idle work optimization in province ground county of table 2 Multiple Time Scales:

Before and after the optimization of the embodiment of the present invention, certain electrical network 500kV main transformer tap joint position is as shown in table 2.By combined optimization, find out the each load tap changer that meets grid net loss minimum.

Large mode Network Loss Rate before and after the integrated idle work optimization in province ground county of table 3 electrical network Multiple Time Scales:

Little mode Network Loss Rate before and after the integrated idle work optimization in province ground county of table 4 electrical network Multiple Time Scales:

According to amount of power supply, Energy loss before and after optimizing under the large mode of Multiple Time Scales province ground county's operational mode idle work optimization uniting and adjustment, by Network Loss Rate computing formula, can calculate the Network Loss Rate of electrical network.Before and after the integrated idle work optimization in province ground county of certain electrical network Multiple Time Scales, large mode Network Loss Rate situation is as shown in table 3, and after optimizing, active power loss is than few 5.89MW before optimizing, and Network Loss Rate reduces by 0.0353%.Before and after the integrated idle work optimization in province ground county of certain electrical network Multiple Time Scales, little mode Network Loss Rate situation is as shown in table 4, and after optimizing, active power loss is than few 4.26MW before optimizing, and Network Loss Rate reduces by 0.0772%.

Should be understood that the present embodiment is only not used in and limits the scope of the invention for the present invention is described.In addition should be understood that those skilled in the art can make various changes or modifications the present invention after having read the content of the present invention's instruction, these equivalent form of values fall within the application's appended claims limited range equally.

Claims (5)

1. the province of the electric power system Multiple Time Scales integrated idle work optimization method in ground county, is characterized in that, comprises the following steps:

(1-1) the electrical network scope that selection will be optimized, obtain the rack data of corresponding electrical network and the service data of electrical network, in computer, be provided with several time scales, user controls computer and selects the rack data of at least one 1 time scale and service data to carry out the trend calculating of province ground county:

(1-1-1) read rack data and service data, rack data comprise Net Frame of Electric Network parameter, and service data comprises the parameter of system operation, set V _i=1 and δ _i=0, set iterations k=0, maximum iteration time k _max=M, generates admittance matrix Y according to parameter of double--layer grids _b; Admittance matrix Y _bin comprise electrical network grid structure relation;

(1-1-2) utilize power flow equation, rated output amount of unbalance Δ P _iwith Δ Q _i:

(1-1-3) verification unbalanced power amount Δ P _iwith Δ Q _ivalue, if max| (Δ P _i, Δ Q _i) | < ε, convergence, then carries out Equivalent Network, wherein convergence precision ε=10 ^-6;

If max| is (Δ P _i, Δ Q _i) |>=ε, utilizes $\left[\begin{array}{cc}\frac{\&PartialD;\mathrm{\&Delta;P}}{\&PartialD;\mathrm{\&delta;}}& \frac{\&PartialD;\mathrm{\&Delta;P}}{\&PartialD;V}\\ \frac{\&PartialD;\mathrm{\&Delta;Q}}{\&PartialD;\mathrm{\&delta;}}& \frac{\&PartialD;\mathrm{\&Delta;Q}}{\&PartialD;V}\end{array}\right]\left[\begin{array}{c}\mathrm{\&Delta;\&delta;}\\ \mathrm{\&Delta;V}\end{array}\right]=-\left[\begin{array}{c}\mathrm{\&Delta;P}\\ \mathrm{\&Delta;Q}\end{array}\right]$ Calculate amount of unbalance Δ V and Δ δ;

(1-1-4) utilize $\left\{\begin{array}{c}\mathrm{\&Delta;V}=V^{(k+1)}-V^{\left(k\right)}\\ \mathrm{\&Delta;\&delta;}={\mathrm{\&delta;}}^{(k+1)}-{\mathrm{\&delta;}}^{\left(k\right)}\end{array}\right.$ Solve V ^(k+1)and δ ^(k+1), make k value increase by 1, as max| (Δ P _i, Δ Q _i) | during>=ε, repeating step (1-1-2);

If (1-2) k > M, the active-power P to network parameter, generator by optimal load flow _gi, reactive power Q _githe size of exerting oneself is adjusted calculating:

(1-2-1) again read rack data and service data, set iterations k ₁=0, maximum iteration time is restricted to k _1max=N, arranges node voltage V _i=1 and node phase angle δ _i=0, slack variable l is set _i=1, u _i=1, Lagrange multiplier z _i=1, w _i=-0.5, y _i=0; σ ∈ (0,1) is called Center Parameter, according to parameter of double--layer grids, generates admittance matrix Y _b;

(1-2-2) work as k ₁during < N, calculate complementary clearance G ap=l ^tz-u ^tw, if complementary clearance G ap < ε, optimal load flow convergence, then carries out Equivalent Network, wherein convergence precision ε=10 ^-6, r is constant;

Wherein, l is a dimensional vector, l ^tthe transposition of l, l=(l ₁, l ₂... l _i), i ∈ r; U is a dimensional vector, u ^tthe transposition of u, u=(u ₁, u ₂... u _i), i ∈ r; Z is a dimensional vector: z=(z ₁, z ₂... z _i), i ∈ r; W is a dimensional vector: w=(w ₁, w ₂... w _i), i ∈ r;

If Gap>=ε, utilizes μ=σ (l ^tz-u ^tw)/(2r) calculation perturbation factor mu, and calculate update equation group $\left\{\begin{array}{c}-[{\&dtri;}_x^{2}f\left(x\right)-{\&dtri;}_x^{2}h\left(x\right)y-{\&dtri;}_x^{2}g\left(x\right)(z+w)]\mathrm{\&Delta;x}+{\&dtri;}_{x}h\left(x\right)\mathrm{\&Delta;y}+{\&dtri;}_{x}g\left(x\right)(\mathrm{\&Delta;z}+\mathrm{\&Delta;w})=L_x\\ {\&dtri;}_x{h}^T\left(x\right)\mathrm{\&Delta;x}=-L_y\\ {\&dtri;}_x{g}^T\left(x\right)\mathrm{\&Delta;x}-\mathrm{\&Delta;l}=-L_z\\ {\&dtri;}_x{g}^T\left(x\right)\mathrm{\&Delta;x}+\mathrm{\&Delta;u}=-L_w\\ \mathrm{Z\&Delta;l}+\mathrm{L\&Delta;z}=-L_l^{\mathrm{\&mu;}}\\ \mathrm{W\&Delta;u}+\mathrm{U\&Delta;w}=-L_u^{\mathrm{\&mu;}}\end{array}\right.,$

$\left\{\begin{array}{c}{L}_x\&equiv;{\&dtri;}_{x}f\left(x\right)-{\&dtri;}_{x}h\left(x\right)y-{\&dtri;}_{x}g\left(x\right)(z+w)=0\\ L_y\&equiv;h\left(x\right)=0\\ L_z\&equiv;g\left(x\right)-l-\underset{\&OverBar;}{g}=0\\ L_w\&equiv;g\left(x\right)+u-\stackrel{\&OverBar;}{g}=0\end{array}\right.;$

(1-2-3) calculate Δ x, Δ y, Δ l, Δ u, Δ z, Δ w, utilize equation $\left\{\begin{array}{c}{\mathrm{\&alpha;}}_p=0.9995\min \{\underset{i}{\min }(\frac{-l_i}{\mathrm{\&Delta;}{l}_i},\mathrm{\&Delta;}{l}_i<0;\frac{-u_i}{\mathrm{\&Delta;}{u}_i},\mathrm{\&Delta;}{u}_i<0),1\}\\ {\mathrm{\&alpha;}}_d=0.9995\min \{\underset{i}{\min }(\frac{-z_i}{\mathrm{\&Delta;}{z}_i},\mathrm{\&Delta;}{z}_i<0;\frac{-w_i}{\mathrm{\&Delta;}{w}_i},\mathrm{\&Delta;}{w}_i>0),1\}\end{array}\right.i=\mathrm{1,2},...,r,$ Calculate iteration step length α _pand α _d;

(1-2-4) utilize $\left\{\begin{array}{c}{x}^{(k+1)}=x^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;x}\\ l^{(k+1)}=l^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;l}\\ u^{(k+1)}=u^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;u}\end{array}\right.$ With $\left\{\begin{array}{c}{y}^{(k+1)}=y^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;y}\\ z_l^{(k+1)}=z^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;z}\\ w^{(k+1)}=w^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;w}\end{array}\right.$ Calculate x ^(k+1), l ^(k+1), u ^(k+1), y ^(k+1),

and w ^(k+1); Make k ₁value increase by 1, proceed to step (1-2-1);

If (1-3) max| (Δ P _i, Δ Q _i) | < ε, Ze Jiangshengdi county data file is carried out Ward Equivalent Network, obtains rack data and system service data;

(1-4) whether repeating step (1-1-1) to (1-1-4) calculating rack data and system service data restrain, if k ₁> N, user controls computer and selects the rack data of yardstick At All Other Times and system service data repeating step (1-1-1) to (1-3) to carry out province ground county trend to calculate;

If (1-5) trend is calculated max| (Δ P _i, Δ Q _i) | < ε parses required rack data and the system service data of the province ground integrated idle work optimization in county of Multiple Time Scales from the equivalent network data file generating;

(1-6) province's required rack data and service data of the ground integrated idle work optimization in county of the Multiple Time Scales parsing carried out to combined optimization:

(1-6-1) read rack data and service data, set iterations k ₂=0, maximum iteration time is restricted to k _max=M2, sets node voltage V _i=1 and node phase angle δ _i=0, set slack variable l _i=1, u _i=1, Lagrange multiplier z _i=1, w _i=-0.5, y _i=0; Wherein, Center Parameter σ ∈ (0,1), generates admittance matrix Y according to parameter of double--layer grids _b;

(1-6-2) work as k ₂during < M2, calculate complementary clearance G ap=l ^tz-u ^tw, if complementary clearance G ap < ε, OPTIMAL REACTIVE POWER is calculated convergence, generates the destination file that comprises each node reactive power distribution of electrical network, transformer station's reactive apparatus switching quantity and load tap changer position, wherein convergence precision ε=10 ^-6;

If Gap>=ε, utilizes μ=σ (l ^tz-u ^tw)/(2r) calculation perturbation factor mu, and calculate update equation group $\left\{\begin{array}{c}-[{\&dtri;}_x^{2}f\left(x\right)-{\&dtri;}_x^{2}h\left(x\right)y-{\&dtri;}_x^{2}g\left(x\right)(z+w)]\mathrm{\&Delta;x}+{\&dtri;}_{x}h\left(x\right)\mathrm{\&Delta;y}+{\&dtri;}_{x}g\left(x\right)(\mathrm{\&Delta;z}+\mathrm{\&Delta;w})=L_x\\ {\&dtri;}_x{h}^T\left(x\right)\mathrm{\&Delta;x}=-L_y\\ {\&dtri;}_x{g}^T\left(x\right)\mathrm{\&Delta;x}-\mathrm{\&Delta;l}=-L_z\\ {\&dtri;}_x{g}^T\left(x\right)\mathrm{\&Delta;x}+\mathrm{\&Delta;u}=-L_w\\ \mathrm{Z\&Delta;l}+\mathrm{L\&Delta;z}=-L_l^{\mathrm{\&mu;}}\\ \mathrm{W\&Delta;u}+\mathrm{U\&Delta;w}=-L_u^{\mathrm{\&mu;}}\end{array}\right.;$

$\left\{\begin{array}{c}{L}_x\&equiv;{\&dtri;}_{x}f\left(x\right)-{\&dtri;}_{x}h\left(x\right)y-{\&dtri;}_{x}g\left(x\right)(z+w)=0\\ L_y\&equiv;h\left(x\right)=0\\ L_z\&equiv;g\left(x\right)-l-\underset{\&OverBar;}{g}=0\\ L_w\&equiv;g\left(x\right)+u-\stackrel{\&OverBar;}{g}=0\end{array}\right.;$

(1-6-3) calculate Δ x, Δ y, Δ l, Δ u, Δ z, Δ w, utilize equation $\left\{\begin{array}{c}{\mathrm{\&alpha;}}_p=0.9995\min \{\underset{i}{\min }(\frac{-l_i}{\mathrm{\&Delta;}{l}_i},\mathrm{\&Delta;}{l}_i<0;\frac{-u_i}{\mathrm{\&Delta;}{u}_i},\mathrm{\&Delta;}{u}_i<0),1\}\\ {\mathrm{\&alpha;}}_d=0.9995\min \{\underset{i}{\min }(\frac{-z_i}{\mathrm{\&Delta;}{z}_i},\mathrm{\&Delta;}{z}_i<0;\frac{-w_i}{\mathrm{\&Delta;}{w}_i},\mathrm{\&Delta;}{w}_i>0),1\}\end{array}\right.i=\mathrm{1,2},...,r,$ Calculate iteration step length α _pand α _d;

(1-6-4) utilize $\left\{\begin{array}{c}{x}^{(k+1)}=x^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;x}\\ l^{(k+1)}=l^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;l}\\ u^{(k+1)}=u^{\left(k\right)}+{\mathrm{\&alpha;}}_p\mathrm{\&Delta;u}\end{array}\right.$ With $\left\{\begin{array}{c}{y}^{(k+1)}=y^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;y}\\ z_l^{(k+1)}=z^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;z}\\ w^{(k+1)}=w^{\left(k\right)}+{\mathrm{\&alpha;}}_d\mathrm{\&Delta;w}\end{array}\right.$ Calculate x ^(k+1), l ^(k+1), u ^(k+1), y ^(k+1),

and w ^(k+1); Make k ₂value increases by 1, proceeds to step (1-6-2);

If (1-7) province of the Multiple Time Scales integrated idle work optimization in ground county calculates complementary clearance G ap < ε, idle work optimization convergence, calculates to the adjusting of main transformer tap changer gear, reactive-load compensator switching amount and part transformer station increase capacitor and reactor details is upgraded initial data file and generate Xin Shengdi county data file according to idle work optimization;

If (1-8) iterations k ₂> M2 means that the integrated idle work optimization in province ground county of Multiple Time Scales calculates the province ground county data file check results not restraining or generate and do not restrain, and adjusts inequality constraints bound, proceeds to step (1-6-1);

(1-9) according to idle work optimization module, calculate the destination file generating after convergence, to the classify electrical network analysis form of formation standard of data;

(1-10) computer is adjusted strategy report according to the analysis result in electrical network analysis form and operational mode, regulate main transformer tap changer gear, regulate each reactive-load compensator switching amount in electrical network and regulate part transformer station in electrical network capacitor and or the quantity of reactor, determine local generator terminal voltage and the power of generator node in the transformer tapping gear, capacitor switching group number, electrical network of Mei Zuo transformer station.

2. the integrated idle work optimization method time scale in the province of electric power system Multiple Time Scales according to claim 1 ground county, is characterized in that, time scale is 5, is respectively year, season, monthly, sky, rolling ultra-short term; Rolling ultra-short term is 15 minutes.

3. the province of the electric power system Multiple Time Scales according to claim 1 integrated idle work optimization method time scale in ground county, it is characterized in that, described rack packet is containing the network line parameter of province Di Xian whole distract or subregion, ground connection branch road parameter and transformer parameter, network line parameter comprises reactance value x _lineij, resistance value r _lineijwith admittance value b _lineij, ground connection branch road parameter comprises electric conductivity value G _iwith susceptance value B _i, transformer parameter comprises reactance value x _lineij, resistance value r _lineijwith no-load voltage ratio k _lineij;

System operational parameters module comprises province's ground whole network in county or the operational factor of subregion and PV node parameter, and operational factor comprises generator node operational factor P _gi, Q _gioperational factor P with load bus _di, Q _di, PV node parameter comprises node voltage V _iwith reactive power bound q _i,

4. the integrated idle work optimization method time scale in the province of electric power system Multiple Time Scales according to claim 1 ground county, is characterized in that, M is 3 to 20, N to be 3 to 20, M2 to be 10 to 100.

5. according to province's integrated idle work optimization method time scale in ground county of the electric power system Multiple Time Scales described in claim 1 or 2 or 3 or 4, it is characterized in that, destination file comprises that the reactive power of whole system distributes, the reactive power compensation amount of each node, transformer voltage ratio tap joint position, reactive power compensator switching amount, node voltage distribute, reactive power exchange amount, the reactive power exchange amount between Di Yu county, the meritorious of power plant between reactive loss, province and the ground of the active loss of system, system are exerted oneself and idle exerting oneself.